氣候變遷下台北地區最大三十分鐘降雨強度推估模式之建立與降雨沖蝕指數之變異點分析

Abstract

最大三十分鐘降雨強度在預測土壤沖蝕十分重要。本研究蒐集1961~2011年之雨量資料，根據Wischmeier and Smith (1978)之定義篩選有效降雨事件，以乘冪函數配合迴歸分析之方式，建立台北地區之最大三十分鐘降雨強度推估模式。模式1僅用雨量推估，模式2則運用雨量及降雨強度因子推估，利用兩種不同模式，藉以探討模擬之準確性。為觀察每十年間之變異性，本研究將1961~2010年之雨量資料，每10年分為一組，共5組，以進行模式係數之運算。經驗證後，確定模式2準確性高且較前人研究之模式快速實用，故將無30分鐘降雨資料的年份(共16年)之各場降雨之降雨量及降雨強度帶入模式2中，可求得台北地區歷年各場I30模擬值。以每年最大之I30代表該年之最大三十分鐘降雨強度情況，由5年移動平均線發現，變化趨勢不穩定，15年移動平均線則是呈現較為穩定、緩步上升之趨勢。證明雖短期變化較不穩定，但以中、長期而言，台北地區的每年最大三十分鐘降雨強度還是呈現穩定增加之趨勢。 在變異點分析中，得知1983年為1961~2011年歷年降雨沖蝕指數時間序列之顯著變異點。台北地區歷年年降雨沖蝕指數的變化趨勢分析，由5年及15年移動平均線顯示，雖短期變化程度較大，但以中、長期而言，係呈現穩定成長之趨勢。從各階段年降雨沖蝕指數統計結果顯示，近年來台北地區的年降雨沖蝕指數隨著氣候變遷之影響，各階段的平均值呈現大幅成長，且變異性顯著增加，其變化趨勢亦符合1983年為降雨沖蝕指數時間序列變異點之結果，綜合觀之，台北地區之年降雨沖蝕指數，已有趨向極端化之現象產生。

The maximum thirty- minute rainfall intensity (I30) is very important in predicting soil erosion. In this study, 51 years of the rainfall data from 1961 to 2011 were collected from Taipei weather station, and effective rainfall events were selected based on the definition by Wischmeier and Smith (1978). The model was established to predict the I30 in Taipei area applying the power function regression. The Model 1 was established only by rainfall factor and the Model 2 was by rainfall and rainfall intensity factor. Examine the accuracy of the simulation by two different models. In order to observe the variability between each decade, the study divided the rainfall data from 1961 to 2010 into 5 groups (every 10 years in one group) to calculate the coefficients of theI30 estimated model. Results proved the Model 2 with high accuracy which was more convenient and more practical than the models of the previous studies. Putting the rainfall and rainfall intensity into the Model 2 can obtain the simulate I30 for 16 years without five minutes rainfall data. Taking annual maximum I30 represent I30. The 5-year moving average trend showed instability, and the 15-year moving average was more stable and rising slowly. Although short-term changes were more unstable, the annual I30 showed a steady increase trend for the long-term in the Taipei area. In the change point analysis, the significant change point of the rainfall erosivity index (the R factor) time series for 1961-2011 is in 1983. The R factor trend displayed by the 5-year and 15-year moving average. The short-period was unstable, but the long-period was showing steady growth trend. Statistical results of the annual R factor shows that the average and variability increased significantly through the impact of climate change. The trend also fit the result of change point analysis (1983 is the change point). Therefore, it is proven that the R factort end to extreme in Taipei area.